Saltar al contenido
Immunogenomics · Neoantigen prediction

Neoantigen prediction and HLA binding

From your somatic variants and the patient’s HLA alleles: peptide–HLA binding prediction and prioritization of candidate neoantigens with an open-source predictor (no reliance on proprietary licenses).

320 €from / sample
Class IHLA-A/B/C
MHCflurryopen-source predictor
Agretopicitymutant vs wild-type
What we do

From mutation to candidate

We turn the tumor’s variants into mutant peptides, predict their binding to the patient’s HLA alleles and rank them by immunogenic potential.

Mutant peptides

Translation of variants (missense and in-frame indels) into peptides of the appropriate length for MHC class I.

8–11 aa

Peptide–HLA binding

Affinity (IC50) and percentile prediction with MHCflurry (class I); classification into binder and strong binder.

IC50 · %rank

HLA typing

We start from the patient’s HLA typing (from sequencing or the one you provide) to restrict the prediction to their alleles.

A/B/C

Prioritization

Ranking by agretopicity (better binding of the mutant than the wild-type) to highlight the most tumor-specific candidates.

log₂(wt/mt)

pVACseq workflow

Complete workflow from the annotated VCF (VEP) + HLA to the table of filtered epitopes, with the predictor configured.

VEP + pVACseq
How it works

Four steps

1. Data

Annotated somatic VCF (VEP) + patient’s HLA alleles.

2. Peptides

Generation of mutant and wild-type peptides.

3. Prediction

MHCflurry for each peptide × patient allele.

4. Prioritization

Ranking by agretopicity and candidate report.

What it includes

Generation of mutant peptides, binding prediction to the patient’s HLA alleles, binder/strong classification, prioritization by agretopicity and report. Volume and academic discounts.

Shall we prioritize your neoantigens?

Send us the annotated somatic VCF and the HLA typing; we get back to you with a quote and timeline within 48 h.

Request a quote See all services
OmicaLabs